Find the Incorrect Data Labeler

Coding Problem: Find the Lowest-Quality Data Labeler You are cleaning a dataset of human annotations. Each record contains: item_id: the ID of the labeled item labeler_id: the ID of the human labeler label: the label assigned by that labeler The same item_id is usually labeled by multiple labelers. Assume the true label of each item is represented by the majority-vote label among all annotations for that item. Find the labeler with the worst overall labeling quality, defined as the labeler_id with the highest fraction of annotations that disagree with the majority-vote label. If multiple labelers have the same error rate, output the lexicographically smallest labeler_id. Input Format The first line contains an integer n, the number of annotation records. The next n lines each contain three strings: item_id labeler_id label Fields are separated by spaces. Output Format Print one string: the labeler_id of the lowest-quality labeler. Constraints 1 <= n <= 200000 item_id, labeler_id, and label are strings without spaces Each item_id has at least one annotation If there is a tie for the majority-vote label of an item, choose the lexicographically smallest label Example Input: 9 q1 alice cat q1 bob cat q1 charlie dog q2 alice dog q2 bob dog q2 charlie cat q3 alice red q3 bob blue q3 charlie red Output: charlie Explanation: For q1, the majority label is cat; charlie is wrong. For q2, the majority label is dog; charlie is wrong. For q3, the majority label is red; bob is wrong. Error rates: alice: 0/3 bob: 1/3 charlie: 2/3 Therefore, output charlie. Example Input 9 q1 alice cat q1 bob cat q1 charlie dog q2 alice dog q2 bob dog q2 charlie cat q3 alice red q3 bob blue q3 charlie red Output charlie